Skip to content
← Notes

Distributed Systems

CAP, consistency, consensus, ordering, and distributed transactions

0/10 chapters · 0 XP earned

  1. 00 Distributed Systems — Roadmap From a single machine to many: failure, time, replication, consensus, and the trade-offs that define every distributed system.
  2. 01 What Makes a System Distributed Why we distribute at all, why independent failure changes everything, and the eight fallacies that trip up every newcomer.
  3. 02 Failure Models & Time How nodes fail, why a network partition is the failure that matters, and why clocks and timeouts cannot be trusted to tell slow from dead.
  4. 03 Replication Keeping copies of data on multiple machines: single-leader, multi-leader, and leaderless designs, sync vs async, and quorums.
  5. 04 Consistency Models What 'consistent' actually means: linearizable, sequential, causal, and eventual consistency, and the trade-offs along the spectrum.
  6. 05 CAP & PACELC The CAP theorem stated precisely, why a partition forces a choice between consistency and availability, and the latency dimension PACELC adds.
  7. 06 Consensus: Raft & Paxos How a group of machines agrees on a single value despite failures: leader election, log replication, and a full walkthrough of Raft.
  8. 07 Ordering & Logical Clocks Ordering events without a shared clock: happens-before, Lamport timestamps, vector clocks, and hybrid logical clocks.
  9. 08 Distributed Transactions Coordinating changes across services: two-phase commit and its blocking problem, sagas, the outbox pattern, and exactly-once myths.
  10. 09 Building Reliable Distributed Systems Putting it together: retries with backoff and jitter, idempotency and deduplication, delivery semantics, and durable execution.